Total expenditure per student in primary education averages USD 13 639 across OECD countries, with most countries in the USD 10 000 to USD 17 000 range (expressed in purchasing power parity (PPP) terms, which adjust for cross-country price differences). There are large differences between the highest-spending countries (over USD 19 000 per student in Iceland, Luxembourg and Norway) and the lowest-spending ones (below USD 5 000 in Colombia and the Republic of Türkiye).
Expenditure is higher at lower secondary level than at primary level, with an OECD average of USD 15 305 per student. This difference is seen in most countries and reflects higher costs linked to subject-specialist teaching, broader curricula and more advanced infrastructure.
Expressing expenditure per student as a share of GDP per capita provides a measure of the economic effort associated with education spending relative to a country’s average income. This measure shows much less cross-country variation than expenditure expressed in USD PPP: most OECD countries cluster between 20% and 25% of GDP per capita at primary level and 21% to 27% at lower secondary level.
Education at a Glance 2026
Context
In all OECD and partner countries, education at the primary and lower secondary levels is compulsory (see Chapter B1). As a result, countries frequently apply similar policies, funding mechanisms and institutional arrangements to both levels, treating them as a unified stage of education. In many OECD countries, primary and lower secondary education are provided within the same structure, often referred to as basic education, reinforcing this integrated approach. Lower secondary education (ISCED 2) is provided as a general programme for all students in most countries, without a division between general and vocational tracks. Since programme orientation becomes more relevant and widespread at the upper secondary level (ISCED 3), issues related to this differentiation are addressed separately in Chapter C4. This chapter examines how funding patterns are changing over time and how resources are allocated within institutions and between different types of institutions.
Direct expenditure on educational institutions, in equivalent USD converted using PPPs for GDP
1. Data include or exclude some categories, please refer to the source table for more information.
2. Year of reference 2022.
For data, see Table C3.1. The data for this figure can be accessed via https://stat.link/e0c4fw.
Other findings
In most OECD countries, private institutions receive much lower levels of government funding than public institutions, although in a few countries either the difference is more moderate (e.g. Belgium and Slovenia, where government expenditure is 10-30% lower per student in private institutions than in public ones) or there is near parity (e.g. Finland).
Education expenditure is strongly driven by staff costs: across OECD countries 91% of total expenditure is dedicated to current expenditure and about 80% of current expenditure is devoted to staff compensation at both primary and lower secondary levels.
Staff compensation per student in primary education increased by around 15% on average in recent years, from USD 7 210 in 2015 to USD 8 035 in 2023, but growth was stronger in countries such as Czechia (USD 3 550 to USD 5 064), the Netherlands (USD 7 250 to USD 9 471) and the Slovak Republic (USD 4 896 to USD 7 204).
Expenditure per student in primary and lower secondary education
Total expenditure per student
Across OECD countries, total expenditure per student in primary education averages USD 13 639 (Figure C3.1). Most countries fall within a range of approximately USD 10 000 to USD 17 000, suggesting broadly comparable spending patterns despite different economic contexts. However, there are pronounced outliers at either end of the scale. High-spending countries such as Luxembourg (USD 31 112), Iceland (USD 21 074) and Norway (USD 19 711) invest much more per student than the OECD average. In contrast, lower expenditure levels are seen in Colombia (USD 3 535) and Türkiye (USD 4 900), as well as in partner and accession countries China, Romania and Peru (below USD 4 500).
At lower secondary level, total expenditure per student tends to be higher than at primary level, with an OECD average of USD 15 305. This is the case in most countries and reflects various factors. At lower secondary level, students are taught by subject specialists (rather than one teacher who covers multiple subjects, as at primary level), which increases staffing needs (more teachers are required per group of students) and involves teachers with higher qualifications and salaries (see Chapter D3). In addition, the curriculum becomes more complex and schools often need to provide facilities such as science labs or IT equipment, which are less common in primary education. Cross-country differences continue to be wide at this level, with Luxembourg (USD 35 223 per student) again at the top end, while Chile (USD 7 725), Colombia (USD 3 603) and Türkiye (USD 5 038) are among the lowest (Figure C3.1)
Differences between public and private institutions
Across OECD countries, reported total expenditure per student (which includes government, private and non-domestic expenditure) is generally higher in public institutions than in private ones. At primary level, the OECD average for total expenditure is USD 14 032 per student in public institutions compared to USD 12 530 in private institutions. At lower secondary level, the difference remains similar, at USD 15 573 in public institutions and USD 14 275 in private institutions (Table C3.1). There are various potential explanations for this pattern, which may seem surprising. One is that institutions are considered private due to the way they are managed, not how they are funded: an institution is classified as private if its overall control and management rest with a non-governmental organisation (OECD, 2018[1]). Within private institutions, further distinction can be made between government-dependent institutions (which rely mostly on government funding) and independent private institutions. The landscape of private institutions and the concentration of different types, as well as their associated funding arrangements vary widely across OECD countries. Another potential reason for lower reported expenditure in private institutions is the difficulty of measuring private expenditure on educational institutions, which may be substantial in some education systems but not fully recorded in official statistics. Household surveys are a common source of data on private expenditure, but depending on the features of the survey, the resulting data may not fully capture the expected categories of household expenditure on education.
There is also considerable variation across countries in the extent of government expenditure on private institutions compared to public ones. In some countries private and public institutions both receive similar amounts of funding per student from government, or the difference is relatively small (up to 15%). For example, in Finland, government spending is almost identical at around USD 13 800 per student in both public and private institutions at primary level. In Sweden private institutions receive slightly lower amounts of government funding per student than public ones, while in Hungary the opposite is the case. In a second set of countries, government expenditure per student is 15-40% lower in private institutions than in public ones. For example, in Slovenia and the United Kingdom, the gap is between 10% and 20% at both primary and lower secondary levels. The gap is slightly larger in Belgium and the Netherlands, where government expenditure per student in private institutions is 20-30% lower than in public ones at both primary and lower secondary levels. Other countries that fall into this group include Australia and Lithuania. Finally, government expenditure per student in private institutions is much lower than in public institutions in about a dozen countries. The gap is near 90% or more in Costa Rica, Greece and Italy, suggesting that in these countries private institutions rely predominantly on private sources of funding. The gap is also relatively wide (over 70%) in Canada, Czechia, New Zealand and Portugal (Table C3.1). In Luxembourg, the results are strongly influenced by the inclusion of the two European schools LUX1 and LUX2. These private schools are mainly financed by the European Union institutions, while government funding accounts for a higher share of funding in other types of private institutions.
Expenditure per student as a share of GDP per capita
Figure C3.2 shows total expenditure per student relative to GDP per capita at primary and lower secondary levels. This indicates the level of resources devoted to each student relative to the average income in a country, providing a measure of the economic effort associated with education spending. This differs from expenditure expressed in USD PPP, which reflects the expenditure per student after adjusting for price differences, but does not indicate how that spending relates to a country’s income level.
Measures of expenditure per student as a share of GDP per capita show a much narrower range than those expressed as USD PPP. At primary level, most OECD countries fall roughly between 16% and 27% of GDP per capita, with a few lower values such as Ireland (10%) and Türkiye (12%), and some higher values such as Korea (28%). At lower secondary level, most countries are between 21% and 30%. Even a country such as Luxembourg, with very high spending expressed in USD PPP terms, appears much closer to the OECD average once expenditure is expressed relative to GDP per capita (20% at primary level and 23% at lower secondary).
Expenditure on educational institutions, in per cent
1. Data include or exclude some categories, please refer to the source table for more information.
2. Year of reference 2022.
For data, see Table C3.5 (web only). The data for this figure can be accessed via https://stat.link/e0c4fw.
Distribution of current expenditure in public institutions
Figure C3.3 shows the share of different types of current expenditure in public institutions (data are not available for all institutions, as the underlying data are not collected for independent private institutions). Across OECD countries, current expenditure in primary education institutions is largely driven by staff compensation. On average, 80% of total current expenditure is devoted to staff compensation, while the remaining 20% is allocated to other current expenditure. Within staff compensation, salaries account for the largest share at 62%, followed by retirement pensions at 11% and other non-salary compensation at 7%. Box C3.1 details how pension contributions are measured or estimated for different types of teachers’ pension schemes.
There is considerable variation across countries in how staff compensation is structured. In some systems, salaries account for a particularly large share of current expenditure. For example, salaries represent 81% of total current expenditure in Chile. Salaries also take up a relatively large share of current expenditure in Austria (72%) and Italy (76%), while pensions and other compensation account for relatively little. In contrast, other countries allocate a larger share to pensions or other types of compensation. In Japan, retirement pensions account for 15% of expenditure, while in Greece they reach 26%. Some systems report relatively high shares of other non-salary staff compensation, which reaches 42% in Costa Rica.
The share of current expenditure allocated to non-staff costs varies significantly as well, with some countries devoting a much larger share to other current expenditure than the OECD average. Finland allocates 40% of its current expenditure to non-staff costs, and Czechia 35%, suggesting relatively higher spending on infrastructure, materials and services. By contrast, Chile (5%), Colombia (6%) and Greece (7%) and dedicate only a small share to these categories.
Direct expenditure within public educational institutions, in per cent
1. Year of reference 2022.
2. Data include or exclude some categories, please refer to the source table for more information.
For data, see Table C3.2. The data for this figure can be accessed via https://stat.link/e0c4fw.
At lower secondary level, the overall structure of current expenditure is very similar to that observed in primary education. On average across OECD countries, 80% of current expenditure is devoted to staff compensation. Within that share, salaries account for 62% of current expenditure, retirement pensions for 11% and other non-salary compensation for 7%. As at primary level, there is a great deal of cross-country variation in the composition of staff compensation. In some systems, salaries account for a particularly large share of expenditure. For example, in Italy salaries represent 87% of current expenditure at lower secondary level, while total staff compensation represents 88%; in Austria, 74% of current expenditure goes on salaries and 75% on staff compensation overall. In contrast, some other countries allocate a larger share of current expenditure to pensions or other compensation. In Colombia, retirement pensions account for 32% of expenditure, and in the United Kingdom they reach 28%. As at primary level, Costa Rica devotes the largest share to other non-salary staff compensation, at 42% of current expenditure, reflecting different compensation structures. The share of expenditure devoted to non-staff costs also varies significantly. Although the OECD average remains 20% of current expenditure, as at primary level, some systems allocate a much larger share to non-staff costs, reaching 40% in Finland and 37% in Czechia. In contrast, Colombia and Ireland report comparatively low shares, at 6 and 7% respectively (Table C3.2, online columns).
Reported expenditure on staff compensation includes not only salaries, but also non-salary staff compensation, and the employer’s cost of pension entitlements accrued by current educational personnel. The features of the pension schemes that apply to teachers affect the way pension contributions are measured and reported in expenditure data.
In fully funded systems, employers make regular contributions to pension funds for current employees, and expenditure corresponds directly to these observed payments. As a result, pension costs are based on actual financial transactions, providing a clear and directly observable measurement of pension contributions linked to current employment.
This applies, for example, in Australia, where teacher pensions are based specific superannuation funds designed for educators. It also partly applies in Canada or the Netherlands, where funded occupational schemes play an important role.
In unfunded or pay-as-you-go (PAYG) systems, employers do not make pension contributions for current employees. Instead, pensions are paid directly from public budgets when retirees claim them. As no current contributions are observed, the cost of pensions must be estimated (imputed) and reported as part of current staff compensation.
In most OECD countries primary and lower secondary teachers’ pensions are largely financed on a PAYG basis. In some countries there is an element of fully funded pension (this is often a smaller component), so that both observed and imputed contributions are needed to measure the full cost of accrued pension rights. For example, in Germany the system is predominantly pay-as-you-go through a statutory pension scheme. A small fully funded component exists in the supplementary pension scheme for public-sector employees in the former East German states. Similarly, in France mixed systems exist but the main system operates on a PAYG basis. In Czechia and Hungary, Italy, Portugal and Spain teachers’ pensions are largely financed on a PAYG basis. A mixed approach exists in Finland, Korea and Sweden, where pension systems combine PAYG components with funded elements and so both observed and imputed contributions are needed to measure the full cost of staff pension entitlements.
Note: This textbox refers to primary and lower secondary education, pension schemes may differ at other education levels.
Trends in expenditure on staff compensation per student
Figure C3.4 shows how expenditure on staff compensation per full-time equivalent student has changed in recent years at primary level. Between 2015 and 2023, staff compensation per student (in constant PPP-adjusted USD) increased on average across OECD countries, from USD 7 210 in 2015 to USD 8 035 in 2023. Several countries recorded particularly strong growth. For example, expenditure on staff compensation per student increased from USD 7 250 to USD 9 471 in the Netherlands, from USD 7 781 to USD 9 220 in Canada and from USD 6 141 to USD 8 054 in Ireland. Significant increases also occurred in Central and Eastern European countries such as Czechia (from USD 3 550 to USD 5 064) and Poland (from USD 5 674 to USD 7 863).
Expenditure in equivalent USD converted using PPPs for GDP, constant prices (2020=100)
1. Data include or exclude some categories, please refer to the source table for more information.
For data, see Table C3.3. The data for this figure can be accessed via https://stat.link/e0c4fw.
Other countries experienced either more moderate changes or little change over the period. Staff compensation per student in the United Kingdom increased slightly at this level from USD 10 419 in 2015 to USD 10 612 in 2023, while there were slight declines in Austria (from USD 9 815 to USD 9 660) and Norway (from USD 10 037 to USD 9 804). Germany also recorded a slight decrease from USD 7 956 to USD 7 888. Declines were more noticeable in a few cases: in Luxembourg expenditure on staff compensation per student fell from USD 19 043 to USD 16 480, although it remains by far the country with the highest expenditure at this level. Japan also recorded a moderate decline, from USD 6 848 to USD 6 503 (Figure C3.4).
These changes have done little to reduce differences in spending on staff compensation per student across OECD and partner countries. High-spending countries such as Luxembourg (USD 16 480), Belgium (USD 11 750), Iceland (USD 11 211) and the United Kingdom (USD 10 612) remain well above the OECD average for primary education in 2023. At the other end of the distribution, countries such as Türkiye (USD 2 613), and accession and partner countries Romania (USD 2 573) and Brazil (USD 2 398) report much lower levels. Although some countries with lower initial levels have increased their expenditure, including Estonia, Hungary and Lithuania, they still remain below the OECD average for this measure (Figure C3.4).
For the definition of expenditure on educational institutions, government expenditure on educational institutions, private expenditure on educational institutions, and public and private educational institutions, refer to Chapter C1.
For an overview of the methodology based on the joint data collection by UNESCO, the OECD and Eurostat (UOE), see Chapter C1. For more detailed information, please refer to the OECD Handbook for Internationally Comparative Education Statistics (OECD, 2018[1]). For country-specific notes, see Education at a Glance 2026 Sources, Methodologies and Technical Notes (https://doi.org/10.1787/dcf64a14-en).
Data refer to the financial year 2023 (unless otherwise specified) and are based on the UNESCO, OECD and Eurostat (UOE) data collection on education statistics administered by the OECD in 2025 (for details see Education at a Glance 2026 Sources, Methodologies and Technical Notes, (https://doi.org/10.1787/dcf64a14-en).
Data for China, India, Indonesia and Saudi Arabia are from the UNESCO Institute of Statistics (UIS), while data for Argentina and South Africa are partly sourced from UIS.
References
[4] European Commission, Eurydice Network (2026), Eurypedia: Description of national education systems, https://eurydice.eacea.ec.europa.eu/eurypedia (accessed on 7 May 2026).
[3] International Organisation of Pension Supervisors (2026), Resources: Pension system profiles of IOPS member jurisdictions/territories, https://www.iopsweb.org/en/resources.html#jurisdiction_profiles (accessed on 7 May 2026).
[2] OECD (2025), Pensions at a Glance 2025: OECD and G20 Indicators, OECD Publishing, Paris, https://doi.org/10.1787/e40274c1-en.
[1] OECD (2018), OECD Handbook for Internationally Comparative Education Statistics 2018: Concepts, Standards, Definitions and Classifications, OECD Publishing, Paris, https://doi.org/10.1787/9789264304444-en.
The data for the figures and tables in this chapter can be downloaded via https://stat.link/e0c4fw.
To access further data and/or other education indicators, please visit the OECD Data Explorer: http://data-explorer.oecd.org/s/4s.
Data cut-off for the print publication 17 June 2026. Please note that the Data Explorer contains the most recent data.
a – category not applicable; b – break in series; d – contains data from another column; m – missing data; u – partial data; x – contained in another column (indicated in brackets). For further control codes, see the Reader’s Guide.
For further methodological information, see Education at a Glance 2026: Sources, Methodologies and Technical Notes, (https://doi.org/10.1787/dcf64a14-en.
Expenditure on educational institutions; in equivalent USD converted using PPPs for GDP; by level of education
1. Includes scholarships and other grants from private sources to households and students.
2. Includes central government expenditure outside educational institutions.
3. Excludes expenditure on private institutions by local governments and international sources.
4. Excludes payments from private sources to educational institutions.
5. Excludes expenditure on vocational lower-secondary institutions.
6. Includes expenditure by non-educational private entities on primary education and payments by households for goods and services required directly or indirectly by educational institutions.
7. Excludes central government expenditure on independent private institutions.
8. Includes government transfers and payments to the non-educational private sector.
9. Excludes expenditure by local and regional governments.
10. Excludes payments by households to private educational institutions and payments by non-educational private entities to independent private institutions.
11. Excludes international expenditure on government-dependent private institutions.
12. Year of reference 2022.
13. Excludes household expenditure and non-domestic expenditure on primary education.
The data for this table can be accessed via https://stat.link/e0c4fw.
In per cent; expenditure from all sources; by level of education
Note: Columns showing data for lower secondary education are available for consultation online.
1. Excludes expenditure on vocational lower-secondary institutions (web columns 13-24).
2. Current expenditure includes expenditure for ancillary services.
3. Year of reference 2022.
The data for this table can be accessed via https://stat.link/e0c4fw.
Expenditure from all sources within public and private educational institutions; constant prices (2020 = 100); by level of education
1. Excludes expenditure on independent private institutions.
2. Excludes expenditure on vocational lower-secondary institutions.
3. Excludes current expenditure on teaching staff in public institutions.
4. Includes expenditure on non-teaching staff in independent private institutions (web columns).
The data for this table can be accessed via https://stat.link/e0c4fw.